Abstract
There is growing evidence that genetic variations in the human leukocyte antigen (HLA) genes play a role in the etiology and clinical course of NHL. HLA-G belongs to the non-classical class I major histocompatibility complex-1 (MHC-I) polymorphic molecules and due to its suppression of immune response it is able to facilitate tumor escape from immunosurveillance. Several polymorphisms in the HLA-G gene have been reported to significantly affect its expression. Therefore, we investigated whether functionally relevant HLA-G polymorphisms, HLA-G-725C/G/T and HLA-G 14-bp, have any influence on the susceptibility to, and clinical course of, diffuse large B-cell lymphoma (DLBCL). The polymorphisms were genotyped in 207 previously untreated patients with DLBCL and 150 unrelated controls. A significant difference in genotype distribution of HLA-G polymorphic genotypes between the patients and controls was found. The frequencies of the HLA-G-725GG or the HLA-G-725GC genotype were lower (OR [odds ratio]= 0.47, P= 0.004), and those of the HLA-G ins/ins genotype were higher (OR= 2.08, P= 0.004) in the patients compared to controls. In univariate logistic regression analysis, neither HLA-G -725C/G/T nor HLA-G 14-bp influenced the probability of achieving a remission. There was no influence of HLA-G polymorphisms on the probability of progression-free survival (PFS). However, the patients carrying the HLA-G-725CC genotype presented a higher probability of 5-year overall survival (OS) than subjects with other genotype combinations of HLA-G-725C/G/T (38.2% vs 21.7%, P=0.003, log-rank test). The estimated 5-year OS among the homozygotes HLA-G del/del was 18.9% compared to the 35.3% in the subjects carrying the HLA-G del/ins or the 46.1% in those with the HLA-G ins/ins genotype (P=0.009, log-rank test). In a multivariate Cox analysis adjusted for IPI factors, we found that both the HLA-G -725C/G/T polymorphism (P= 0.01) and the IPI (P< 0.0001) retained their independent prognostic impact on OS The influence of the particular genotypes of the HLA-G -725C/G/T and the HLA-G 14-bp polymorphism on OS allow us to single out two HLA-G genotype-based risk groups. The low risk (LR) group included the patients carrying the HLA-G -725CC genotype and the HLA-G del/ins or the HLA-G ins/ins genotype. In contrast, the high risk (HR) group comprised those patients with other HLA-G genotype combinations. It is worth noting that the estimated 5-year OS rate of patients with LR genotypes was 42.7% in comparison to the 19.3% (P= 0.001, log-rank test) in patients with HR genotypes. An additional multivariate analysis, including HLA-G genotype-based risk groups and the IPI, revealed that both the intermediate high/high IPI risk group (P< 0.0001) and the HR genotype group (P= 0.004) independently increased the risk of death.
This is the first study indicating an important role for HLA-G polymorphisms in the clinical course of DLBCL. The potential influence of HLA-G polymorphisms on the susceptibility to DLBCL deserves further study. It would seem that the inherited ability of the host to suppress anti-tumor immune response and then facilitate tumor escape from immunosurveillance might contribute to the pathogenesis and prognosis of B-cell malignancies.
Robak:MorphoSys AG: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.